NO STATISTICS NEED APPLY: Half the statistics disappear!

THURSDAY, APRIL 2, 2015

Part 4—“Reporting” the nation's schools: Just as a brief aside, eleven “educators” are in prison today in Atlanta after a six-month trial

A six-month trial!

Don’t get us wrong! The conduct is which these teachers engaged was ludicrous, inane and baldly dishonest.

That said, we think this prosecution was giant overkill. In this Lenten season, you might even say that these teachers have died for the New York Times’ (statistical) sins.

These teachers were convicted of ludicrous cheating on Georgia’s statewide standardized tests. They even played the erasure game, systematically changing wrong answers to right after test sessions were over.

This is deeply ludicrous conduct, but it had been going on for decades by the time these teachers got caught. In large part, the practice had continued apace because of the incompetence (and/or the dishonesty) of the national “press corps.”

We have a personal history here. Quickly, let’s review it:

We first reported outright cheating to the Baltimore Sun in the early 1970s. In 1981, we published this op-ed column on the subject in the Baltimore Evening Sun.

By that time that column appeared, we had learned something which we found startling at the time:

Even back then, publishers of the big standardized tests would, for an extra fee, electronically scan a school system’s answer sheets looking for the erasure patterns which indicate widespread cheating.

More than thirty years ago, this phenomenon was well understood! (We were told about this practice by a high official at one of the large test batteries, but it wasn't a secret.)

It wasn’t just the fact that the big test publishers were providing that remarkable service. On a basic statistical basis, the scoring patterns at certain schools were clownishly obvious. Except to the nation’s “press corps,” which kept failing to report this ongoing phenomenon.

Inside the press corps, statistics are hard! Also, preferred narratives are powerful.

A bit more personal history:

In the late 1980s, we consulted with Dr. John Cannell about this general topic. He achieved ten minutes of fame by suggesting, through reference to his humorous Lake Wobegon Effect, that this sort of cheating may have been widespread at that time.

Cannell’s suggestions were widely reported, for roughly ten minutes. The topic was then completely abandoned—but wait, there’s more:

In 2006, we discovered a clownish statistical scam which was elevating “passing rates” at schools all over the state of Virginia. Quite forthrightly, the chairman of the state board acknowledged to us, in an on the record interview, that this remarkable conduct actually had occurred.

For background, just click here, with a link to earlier reports.

The Washington Post, and other Virginia newspapers, failed to report this news. A rather remarkable statewide scam went almost wholly unreported.

One more bit of personal history:

In 2007, a little-known person named Michelle Rhee was selected to become the head of the Washington D.C. schools. She arrived in Washington trailing dramatic claims about the sensational test scores she had produced in her three years as a Baltimore second and third grade teacher.

As we noted in real time, the test scores Rhee said her students recorded seemed absurd on their face. The Washington Post averted its gaze. A few years later, a cheating scandal matching Atlanta’s surfaced in D.C.’s schools.

Let’s give credit where credit is due! In this morning’s New York Times, Alan Blinder wrote the full-length news report about the Atlanta convictions.

In the hard-copy Times, Blinder’s report leads the National section. He offers this bit of history about the Atlanta case:

BLINDER (4/2/15): In 2009, The Atlanta Journal-Constitution started publishing a series of articles that sowed suspicion about the veracity of the test scores, and Gov. Sonny Perdue ultimately ordered an investigation.

The inquiry, which was completed in 2011, led to findings that were startling and unsparing: Investigators concluded that cheating had occurred in at least 44 schools and that the district had been troubled by “organized and systemic misconduct.” Nearly 180 employees, including 38 principals, were accused of wrongdoing as part of an effort to inflate test scores and misrepresent the achievement of Atlanta’s students and schools.

Decades after this basic problem was obvious and widely known, reporters at the Journal-Constitution actually did some reporting!

That paper deserves a lot of credit for its work. So does USA Today, which did the basic reporting on D.C.’s testing debacle.

Those newspapers deserve full credit for some belated work. But by the time of the Journal-Constitution’s reports, its findings could only seem “startling” to completely clueless reporters like the world-class stumblebums who work for the New York Times.

How is it possible? How could big newspapers like the Post and the Times remain so clueless about this matter for so long? Before considering last week’s work on the nation's schools, we’ll offer two possibilities:

Basic incompetence: Many “journalists” at these newspapers are almost completely incompetent. They lack even the most basic skills for processing even the simplest statistics. Last week’s report about the gender income gap among nurses represents a crowning example of a phenomenon which is widespread.

Devotion to narrative: These newspapers also tend to defer to the narratives which shape so much of our news. They’ve always loved ridiculous stories about miraculous test scores at low-income schools. These types of stories maintained preferred beliefs about the educational challenges we face and the reasons for the nation’s “achievement gaps.” At the same time, they never seemed to have a taste for reporting which may have undermined traditional, Dick-and-Jane era faith in the inerrant probity of teachers.

By today, preferred elite narratives about public schools are different, and so is the press corps’ clowning. Consider several events last week which had our youthful analysts gnashing what’s left of their teeth.

First, a small appetizer. On Sunday, March 22, Martin and Bosman authored a front-page report in the New York Times about Chicago’s mayoral race. Will anything ever stop our scribes from typing variants of the highlighted standard statement?

MARTIN AND BOSMAN (3/22/15): Another challenge for hard-line progressives is that Mr. Emanuel, while heavily funded by business interests and some wealthy Republicans, has a record of accomplishment that covers many liberal priorities. He may have angered the teachers union and many black residents by closing 50 failing schools in mostly minority neighborhoods. But he passed full-day kindergarten, a move similar to Mr. de Blasio’s expansion of prekindergarten. Mr. Emanuel also added free community college for high-achieving public school graduates...

The highlighted statement can be defended as “technically accurate.” Emanuel did close fifty schools; this did anger some residents. It’s also true that the schools in question were mostly in minority neighborhoods or, in a softer formulation, were “in mostly minority neighborhoods.”

That said, Chicago’s student population is 90 percent black and Hispanic. There's no way to close underutilized schools in Chicago without closing schools “in mostly minority neighborhoods.”

This situation obtains in urban systems all over the country. But at various major newspapers, reporters insist on feigning surprise at the fact that underutilized schools are getting closed in black and Hispanic neighborhoods.

These ace reporters seem immune to the simplest statistical observations. That silly script-read in the Times was just the latest example.

That was a minor if nagging example. Last Sunday, March 29, Fareed Zakaria swung from the heels in the featured front-page piece in the Washington Post’s Outlook section.

Zakaria painted a Mandated Thoroughly Standard Press Portrait of the state of the nation’s schools. “Every month, it seems, we hear about our children’s bad test scores in math and science,” he said as he started, understating the ubiquity of this mandated portrait.

As he continued, he made no attempt to challenge this gloomy portrait. Indeed, referring to scores attained by American students on international tests, he reinforced the mandated picture:

“My point is not that it’s good that American students fare poorly on these tests. It isn’t,” the serial plagiarist said.

(For the record, there’s no obvious way for anyone to cheat on these international tests, nor is there any apparent motive.)

As he continued, Zakaria reacted to his gloomy portrait in a surprising way. He said it doesn’t hugely matter if American kids score poorly on international tests.

You can read his piece to evaluate his reasoning. Here’s what had our youthful analysts gnashing their teeth:

Like everyone else is America’s press corps, Zakaria engaged in a bit of statistical sleight-of-hand. As everyone except our “reporters” knows, there are two major international test batteries in which most developed nations participate—the somewhat newer PISA, on which American kids score somewhat less well, and the somewhat older TIMSS/PIRLS, on which American kids score somewhat better.

Like everyone else in America’s “press corps,” Zakaria only cited scores from the PISA. Quite literally, he discarded half the data in reaching his Mandated Thoroughly Standard Gloomy Press Corps Portrait of These Kids Today!

Who except American “journalists” function this way? Who else conducts an analysis by ignoring half the data?

In the modern context, powerful figures seem to want this topic reported in the gloomiest possible way. Within the upper-end press, almost everyone does.

In truth, they’re all serial plagiarists! For reasons only they can explain, they’re typing the insider line, in which American kids are a pitiful mess thanks to the sloth of their unionized teachers.

Down in Atlanta, eleven teachers are in prison today. They died for the sins of the New York Times, which spent decades failing to report the fact that cheating was going on in some parts of the casino.

Today, our journalists keep disappearing the TIMSS. When it comes to test scores on the NAEP, they report the gaps but disappear the large gains, as we’ve explained a million times by now.

They keep discarding half the news! It’s how these life forms roll.

Tomorrow: The latest statistics on campus